Trusting AI in Production: Building Observability for AI Agents at Scale

How top teams monitor AI agents in production—using telemetry and observability to catch issues early, prevent drift, and scale with confide

In this webinar, Monte Carlo and Snowflake come together to explore what it actually takes to operate agentic AI systems with confidence. We'll look at how leading teams are using native telemetry alongside production monitoring to catch problems early, maintain reliability at scale, and build the kind of trust in their AI systems that lets them move faster over time.

You’ll learn:

1️⃣ How mature AI teams think about the lifecycle of an agent after it ships — and what separates teams that react to failures from teams that prevent them.

2️⃣ What production-grade observability looks like in practice — including a hands-on look at how teams using Snowflake Cortex Agents are leveraging native telemetry to monitor performance, track token consumption, and catch degradation before it becomes a business problem.

3️⃣ The early warning signals that indicate an agent is drifting — and a practical framework for acting on them before something downstream breaks.